A major challenge in genetics is to identify genetic variants driving phenotypic variation, and the analysis of natural variation in human protein sequences is an important avenue to meet the challenge. As weak noncovalent interactions play critical role in protein folding, assembly and recognition, structural analysis of how noncovalent interactions in human proteins, from one individual to another, vary will be critical. Our results suggest that thousands of noncovalent interactions, particularly weak ones (e.g., p-interactions, anion-quadrupole (AQ), hydrogen bonds etc.), are perturbed in human proteins due to natural variation. Like other p-interactions, AQ also plays important role in macromolecular structure. However, the strength of weak interactions (e.g., AQ) remain poorly understood, and therefore, the interpretation of the consequences of natural variation of human protein sequences remain incomprehensible. The absence of the knowledge of weak-interaction energetics and the comprehensive map of all weak interactions altered in human proteins will continue to significantly contribute to the lack of understanding of the origin of phenotypic variation. Continued existence of this knowledge gap represents an important problem because, until it is filled, how genetic variants drive phenotypic variation remain incomprehensible for beneficial genetic interventions. Our long-term goal is to better understand the role of weak noncovalent interactions in regulating protein function. The objective for this particular R15 application is to comprehensively measure the strength of a weak interaction (e.g., AQ) and to create a comprehensive catalogue of all noncovalent interaction in human protein structures that are altered due to natural human sequence variation. Our rationale is that (a) determination of the energy of a weak interaction (e.g., AQ) is likely to provide new insights by enabling subsequent studies on protein function by manipulating AQ; (b) the availability of a complete catalogue of fine structural details of natural missense variants of all human proteins will facilitate probing molecular mechanism of genetic variants driving phenotypic variation. The two specific aims are: 1) Determine the strength of AQ experimentally; 2) Create a database of 3D structural maps of natural missense variants of human proteins. For Aim-1, using 18 carefully chosen protein-peptide interfaces, we experimentally measure the strength of AQ that occur in various structural contexts for a comprehensive estimate. For Aim-2, using human exome aggregation consortium (ExAC) database, we provide a comprehensive, fine structural map of all noncovalent interactions in human protein structures that are perturbed due to natural variation. Using molecular dynamic simulations, we also probe the consequences of ExAC mutations in two functionally important human proteins. The approach is innovative for capturing a link between genetic and phenotypic variations at atomic resolution. The research is significant, because it is expected to vertically expand the understanding of how genetic variations contribute to phenotypic variation. That will enable preventative, therapeutic manipulations of human proteome.